Infant feeding clusters are associated with respiratory health and allergy at school age in the PARIS birth cohort

International audience; BACKGROUND: As infant feeding may influence allergy development, we aimed to identify groups of infants based on feeding practices and to examine their associations with respiratory health/allergy at 8 years in the PARIS birth cohort. METHODS: Data on breastfeeding, consumption of infant formula (regular, pre-/probiotics, partially hydrolysed with hypoallergenic label [pHF-HA], extensively hydrolysed [eHF], soya) and solid food introduction were collected using repeated questionnaires at 1, 3, 6, 9 and 12 months. Infants with similar feeding practices over the first year of life were grouped using multidimensional longitudinal cluster analysis. Respiratory/allergic morbidity was studied at 8 years as symptoms, doctor's diagnoses (asthma, hay fever, eczema, food allergy), and measurement of lung function, FeNO and specific IgE. Associations between feeding-related clusters and respiratory/allergic morbidity were investigated using multivariable logistic and linear regression models adjusted for potential confounders including early respiratory/allergic outcomes and parental history of allergy. RESULTS: Five clusters were identified among 3,446 infants: Cluster 1 (45%) mainly fed with regular formula, Cluster 2 (27%) exclusively breastfed during the first 3 months, and three other clusters consuming different types of formula (pre-/probiotics for Cluster 3 [17%], pHF-HA for Cluster 4 [7%], eHF/soya for Cluster 5 [4%]). Compared to Cluster 1, children from Cluster 2 tended to have a lower risk of asthma and children from Cluster 4 had a significant lower lung function (FEV1 , FVC), higher FeNO and higher risk of sensitisation at 8 years. CONCLUSION: Early pHF-HA use was negatively associated with objective measures of respiratory/allergic morbidity at school age, while children breastfed for at least 3 months seem protected against asthma at 8 years old.